You are on page 1of 3

ARTICLE IN PRESS

Personality and Individual Differences xxx (2008) xxxxxx

Contents lists available at ScienceDirect

Personality and Individual Differences


journal homepage: www.elsevier.com/locate/paid

Short Communication

Caffeine, stress, and proneness to psychosis-like experiences:


A preliminary investigation
Simon R. Jones *, Charles Fernyhough
Department of Psychology, Durham University, South Road, Durham DH1 3LE, UK

a r t i c l e

i n f o

Article history:
Received 8 July 2008
Received in revised form 15 October 2008
Accepted 31 October 2008
Available online xxxx
Keywords:
Coffee
Hallucination
Persecutory ideation
Psychosis
Schizophrenia
Tea

a b s t r a c t
In diathesisstress models of psychosis, cortisol released in response to stressors is proposed to play a
role in the development of psychotic experiences. Individual differences in cortisol response to stressors
are therefore likely to play a role in proneness to psychotic experiences. As caffeine has been found to
increase cortisol response to a given stressor, we proposed that, when levels of stress were controlled
for, caffeine intake would be related to hallucination-proneness and persecutory ideation. Caffeine intake,
stress, hallucination-proneness and persecutory ideation were assessed by self-report questionnaires in a
non-clinical sample (N = 219). Caffeine intake was positively related to stress levels and hallucinationproneness, but not persecutory ideation. When stress levels were controlled for, caffeine intake predicted
levels of hallucination-proneness but not persecutory ideation. Implications of these findings are discussed and avenues for future research suggested.
2008 Published by Elsevier Ltd.

1. Introduction
Diathesisstress models of psychosis propose that stress may
play a contributory role in the development of hallucinations and
delusions, typically associated with the schizotypal/schizophrenia
spectrum (Neuchterlein & Dawson, 1986). Cortisol has been highlighted as having a key role in translating the experience of psychosocial stressors into the biological factors associated with
psychosis (Walker & Diforio, 1997). It therefore seems likely that
factors causing individual differences in cortisol response to stressors will influence the likelihood of developing hallucinatory or
delusional experiences. One such factor is caffeine, which has been
found to cause an increase in the amount of cortisol released in response to a stressor (Lane, Adcock, Williams, & Kuhn, 1990).
Although daily usage of caffeine creates a reduction in this effect,
caffeine intake in habitual users still enhances the cortisol response to stress (Lovallo et al., 2005).
As such, caffeine intake may be associated with greater proneness to psychotic experiences due to its enhancement of the cortisol response to stress. We hence firstly hypothesised that, for a
given level of stress, caffeine intake would be associated with
greater levels of proneness to psychotic experiences. Controlling
for stress levels is also necessary due to caffeine consumption
increasing when individuals are stressed (Ratliff-Crain & Kane,
1995). We secondly hypothesised that caffeine intake would inter* Corresponding author. Tel.: +44 1913 343240; fax: +44 1913 343241.
E-mail address: s.r.jones@durham.ac.uk (S.R. Jones).

act with stress, with caffeine intake being more strongly associated
with proneness to psychotic experiences when levels of stress
were high. No studies have yet tested such a hypothesis.
Existing studies of the relation between caffeine and the positive symptoms of schizophrenia (e.g., clinically relevant hallucinations and delusions) have produced mixed findings. Studies that
have experimentally altered caffeine intake in patients with
schizophrenia, by switching patients to decaffeinated coffee for a
period of weeks, show conflicting results. Whilst some have found
resultant decreases in positive symptoms (e.g., De Freitas & Schwartz, 1979), other studies have failed to replicate this (e.g., Mayo,
Falkowski, & Jones, 1993). In the only study to systematically
administer caffeine (10 mg/kg) directly to patients with schizophrenia, Lucas et al. (1990) found that suspiciousness/paranoia
did not change, but that grandiosity and unusual thought content
increased (data on hallucinations were not reported). However,
such studies may be confounded by the interaction of caffeine with
antipsychotic medication (Haslemo, Eikeseth, Tanum, Molden, &
Refsum, 2006).
One way to avoid this confound is to assume that psychotic
experiences exist on a continuum stretching into the healthy population (Johns & van Os, 2001) and examine the relation between
caffeine and psychosis-like experiences in the general population.
The only study to date to examine the relation between caffeine intake and schizotypy (Larrison, Briand, & Sereno, 1999) found no
relation between the two. However, this study used a crude self-report measure of caffeine intake in which participants were simply
asked to circle the number of caffeinated. . .drinks they consumed

0191-8869/$ - see front matter 2008 Published by Elsevier Ltd.


doi:10.1016/j.paid.2008.10.032

Please cite this article in press as: Jones, S. R., & Fernyhough, C. Caffeine, stress, and proneness to psychosis-like experiences: ... Personality
and Individual Differences (2008), doi:10.1016/j.paid.2008.10.032

ARTICLE IN PRESS
2

S.R. Jones, C. Fernyhough / Personality and Individual Differences xxx (2008) xxxxxx

per day or week (p. 103), and then split individuals into high,
moderate and low intake groups, hence reducing the power of
the study. Furthermore, it employed a general measure of schizotypy, with no focus on specific experiences such as hallucinations
or feelings of persecution. We accordingly set out to test our
hypotheses in a non-clinical population, with a methodology chosen to avoid the limitations of the Larrison et al. study.

have been under during the last year. Examples of items include
You feel tense and You have many worries. Items are scored
on a four-point Likerts scale ranging from Almost Never (1) to
Usually (4). A total score is derived using the formula ((raw score
30)/90).

2. Method

Cronbachs alphas for all scales employed were greater than 0.7,
and hence satisfactory. Caffeine content for DCI items was calculated using the values presented in Table 1. The caffeine content of
novel energy drinks and caffeine tablets was determined by reference to the relevant manufacturers website. Descriptive statistics
are presented in Table 2. Mean daily caffeine consumption of
141 mg was comparable to previous studies in student populations
(e.g., Landrum, 1992). Bivariate non-parametric correlational analyses (using a Bonferroni corrected alpha of .05/3, a = .02) showed caffeine intake/kg to correlate significantly with stress, q = .17, p < .02,
and LSHS-M, q = .22, p < .02, but not PIQ scores, q = .08, n.s. PIQ and
LSHS-M scores were also significantly correlated, q = .40, p < .001.
Parametric statistical analysis was then performed using two
hierarchical multiple linear regressions (MLRs) with LSHS-M and
PIQ score as the respective dependent variables. Age, gender, and
stress were entered in the first step, with caffeine/kg and the
(mean-centered) interaction between stress and caffeine/kg entered in a second step. Results are presented in Table 3. Residuals
were normally distributed (D = .05. p > .05, for both MLRs) and
there was no evidence of multicollinearity in the data. In the first
step, stress was found to predict LSHS-M and PIQ scores. After
age, gender and stress levels had been controlled for in the first
step, the second step was found to explain a significant amount
of variance in LSHS-M but not PIQ scores (b values for the second
step of the MLR of PIQ scores are hence not reported). Caffeine/
kg predicted LSHS-M but not PIQ scores. The interaction between
stress and caffeine/kg was not a significant predictor of either
LSHS-M or PIQ scores.

2.1. Participants
Students (N = 219, 154 women) at a United Kingdom university,
with a mean age of 20.1 years (SD = 1.34, range = 1828) were recruited through e-mail invitation. No financial incentive was offered and the likelihood of repeated participation was thus
considered low. Answers were given anonymously, with only the
age, gender and weight of the participants being recorded. Cigarette smokers were excluded from the study.
2.2. Measures
Participants completed the following questionnaires in the order stated below:
Durham Caffeine Inventory (DCI): Caffeine was assessed utilising
a new tool designed to assess caffeine intake in a contemporary UK
student population. The DCI presents specific types of drink and
food containing caffeine (see Table 1), of which participants rate
their typical intake over the past year, using a 12-point response
scale ranging from none/less than one per week to 8+ per day.
In addition to predefined categories (e.g., instant coffee, tea) participants are also asked about coffee purchased from coffee shops,
other energy drinks, and caffeine tablets. Participants use the same
response scale, but are free to enter the name of the beverage or
make of tablets used. A full copy of the DCI is available upon
request.
Hallucination-proneness: This was assessed using the modified
Launay-Slade Hallucination Scale (LSHS-M: Laroi & van der Linden,
2005), a 16-item instrument designed to measure predisposition to
hallucination-like experiences. Each item is scored on a five-point
Likerts scale ranging from Certainly applies to me (4) to Certainly does not apply to me (0).
Proneness to persecutory delusion-like beliefs: The Persecutory
Ideation Questionnaire (PIQ; McKay, Langdon, & Coltheart, 2006)
is a 10-item questionnaire designed to measure persecutory ideation in both clinical and non-clinical samples. Items are rated on
a five-point Likerts scale ranging from Very True (4) to Very Untrue (0).
Stress: This was determined by the Perceived Stress Questionnaire (Levenstein et al., 1993), a 30-item instrument which assesses the levels of stress an individual perceives themselves to

Table 1
Caffeine data utilized.
Beverage

Caffeine (mg)

Source

Tea (weak)
Tea (medium)
Tea (strong)
Coffee (instant)
Coffee (brewed)
Coffee (e.g., Starbucks)
Cola-drinks
Red Bull
Solid milk chocolate
Solid dark chocolate
Caffeine tablets

25
42
51
45
111
188
35
80
6
20
100

Food Standards Agency (2004)


FSA (2004)
FSA (2004)
FSA (2004)
FSA (2004)
McCusker, Goldberger, and Cone (2003)
(details from the authors)
Alford, Cox, and Wescott (2001)
(details from the authors)
(details from the authors)
Manufacturer website

3. Results

Table 2
Descriptive statistics for variables under investigation.

Caffeine intake (mg/day)


Caffeine intake (mg/day/kg)
LSHS-M
PIQ
Stress

Mean (SD)

Range

143.48 (167.22)
2.31 (2.88)
23.81 (12.18)
8.71 (6.87)
.45 (.17)

1951
.0120.15
163
033
.101.00

Note: LSHS-M = modified Launay-Slade Hallucination Scale; PIQ = Persecutory Ideation Questionnaire.

Table 3
Multiple linear regression analysis.
LSHS-M

PIQ

R2 (age, gender, stress)


F(3,214)
b (age)
b (gender)
b (stress)

.18
15.77, p < .001
.11, n.s.
.01, n.s.
.42, p < .001

.28
28.07, p < .001
.06, n.s.
.20, p < .001
.53, p < .001

R2 (age, gender, stress, caffeine/kg,


stress*caffeine/kg)

.21

.29

F(5,212)
DF(2,212)
b (age)
b (gender)
b (stress)
b (caffeine/kg)
b (stress*caffeine/kg)

11.29, p < .001


3.89, p < .05
.10, n.s.
.01, n.s.
.39, p < .001
.18, p < .01
.04, n.s.

17.48, p < .001


1.44, n.s.

Please cite this article in press as: Jones, S. R., & Fernyhough, C. Caffeine, stress, and proneness to psychosis-like experiences: ... Personality
and Individual Differences (2008), doi:10.1016/j.paid.2008.10.032

ARTICLE IN PRESS
S.R. Jones, C. Fernyhough / Personality and Individual Differences xxx (2008) xxxxxx

4. Discussion
The present study set out to examine the relations among caffeine intake, stress and proneness to psychosis-like experiences
in a non-clinical population. Correlational analyses showed that
caffeine intake was positively related to stress levels. This was in
line with previous research (Ratliff-Crain & Kane, 1995). Caffeine
intake was correlated with hallucination-proneness, but not persecutory ideation. The latter finding is consistent with the work of
Lucas et al. (1990), which found that caffeine administration (in
patients with schizophrenia) did not lead to increases in suspiciousness/paranoia.
When multiple linear regression analyses were performed, controlling for stress, caffeine intake per kg predicted hallucinationproneness but not persecutory ideation. The present study hence
offers some support for our first hypothesis, namely that when levels of stress are controlled for, caffeine intake is positively related
to levels of psychosis-like experiences. However, the effect was
found to be weak and specific to hallucination-proneness, and
not persecutory ideation. Our second hypothesis, that the interaction between caffeine intake and stress would predict hallucination-proneness and persecutory ideation, was not supported.
A number of limitations of this study need to be acknowledged.
Firstly, we cannot eliminate the possibility that hallucinationproneness could be a cause rather than a result of increased caffeine intake. This would be consistent with the finding that caffeine
intake can act as a coping mechanism to bring relief from problems
(Ratliff-Crain & Kane, 1995). Secondly, the present study utilised a
non-clinical sample, and it is unclear if the present results are
likely to be generalizable to clinical populations. Finally, we used
a retrospective self-report measure of caffeine that was not independently validated.
Our findings suggest a number of fruitful avenues for future research. Firstly, research may wish to address whether the relation
found here between hallucination-proneness and caffeine intake is
causal. For example, caffeine tablets could be administered to participants before undertaking signal detection tasks to determine if
caffeine intake does lead to increased levels of hallucinatory experience. Secondly, research may wish to address the mechanisms
which may lead to caffeine causing hallucinatory experiences.
One interpretation of our finding of a relation between caffeine
and hallucination-proneness, but not persecutory ideation, is that

the former relation exists not because of potentiation of the cortisol response to stressors (which would presumably also lead to a
link with persecutory ideation), but as a result of other mechanisms. It would also be profitable to revisit studies of caffeine intake in patients with schizophrenia. The present study offers
preliminary evidence that, if levels of stress are controlled, caffeine
intake in patients may relate to levels of hallucinatory experiences.
References
Alford, C., Cox, H., & Wescott, R. (2001). The effects of Red Bull Energy Drink on
human performance and mood. Amino Acids, 21, 139150.
De Freitas, B., & Schwartz, G. (1979). Effects of caffeine in chronic psychiatric
patients. American Journal of Psychiatry, 136, 13371338.
Food Standards Agency (2004). Survey of caffeine levels in hot beverages: Food
surveys. United Kingdom: Food Standards Agency.
Haslemo, T., Eikeseth, P. H., Tanum, L., Molden, E., & Refsum, H. (2006). The effect of
variable cigarette consumption on the interaction with clozapine and
olanzapine. European Journal of Clinical Pharmacology, 62, 10491053.
Johns, L. C., & van Os, J. (2001). The continuity of psychotic experiences in the
general populations. Clinical Psychology Review, 21, 11251141.
Landrum, R. E. (1992). College students use of caffeine and relation to personality.
College Student Journal, 26, 151155.
Lane, J. D., Adcock, R. A., Williams, R. B., & Kuhn, C. M. (1990). Caffeine effects on
cardiovascular and neuroendocrine responses to acute psychosocial stress and
their relationship to level of habitual caffeine consumption. Psychosomatic
Medicine, 52, 320336.
Laroi, F., & van der Linden, M. (2005). Nonclinical participants reports of
hallucinatory experiences. Canadian Journal of Behavioural Science, 37, 3343.
Larrison, A. L., Briand, K. A., & Sereno, A. B. (1999). Nicotine, caffeine, alcohol and
schizotypy. Personality and Individual Differences, 27, 101108.
Levenstein, S., Prantera, C., Varvo, V., Scribano, M. L., Berto, E., Luzi, C., et al. (1993).
Development of the perceived stress questionnaire a new tool for
psychosomatic research. Journal of Psychosomatic Research, 37, 1932.
Lovallo, W. R., Whitsett, T. L., alAbsi, M., Sung, B. H., Vincent, A. S., & Wilson, M. F.
(2005). Caffeine stimulation of cortisol secretion across the waking hours in
relation to caffeine intake levels. Psychosomatic Medicine, 67, 734739.
Lucas, P. B., Pickar, D., Kelsoe, J., Rapaport, M., Pato, C., & Hommer, D. (1990). Effects
of the acute administration of caffeine in patients with schizophrenia. Biological
Psychiatry, 28, 3540.
Mayo, K. M., Falkowski, W., & Jones, C. A. (1993). Caffeine: Use and effects in longstay psychiatric patients. British Journal of Psychiatry, 162, 543545.
McCusker, R. R., Goldberger, B. A., & Cone, E. J. (2003). Caffeine content of speciality
coffees. Journal of Analytical Toxicology, 27, 520522.
McKay, R., Langdon, R., & Coltheart, M. (2006). The persecutory ideation
questionnaire. Journal of Nervous and Mental Disease, 194, 628631.
Neuchterlein, K. H., & Dawson, M. E. (1986). A heuristic vulnerability/stress mode of
schizophrenic episodes. Schizophrenia Bulletin, 10, 300312.
Ratliff-Crain, J., & Kane, J. (1995). Predictors for altering caffeine consumption
during stress. Addictive Behaviors, 20, 509516.
Walker, E. F., & Diforio, D. (1997). Schizophrenia: A neural diathesis-stress model.
Psychological Review, 104, 667685.

Please cite this article in press as: Jones, S. R., & Fernyhough, C. Caffeine, stress, and proneness to psychosis-like experiences: ... Personality
and Individual Differences (2008), doi:10.1016/j.paid.2008.10.032